After crossing the barriers put up by nominal, or real, stickiness’s, inflation remains a monetary (from whatever source) phenomenon! But many doubt that simple “truth”. In 1986 the late Rudi Dornbusch, for example, wrote: “The impact of the exchange rate on inflation is a well established fact for any banana republic and even for industrialized countries…”.
Recent history is always at the forefront of our minds and the “visible” history of the dollar exchange rate against major currencies and inflation, to Dornbusch and others writing at the time, is described in figure 1. It was “natural”, then, to expect that since the dollar began depreciating in early 1985, that inflation would soon pick up. It didn´t.
On August 15th 1971, President Nixon closed the “gold window” and decreed a 90-day price-wage freeze fearful that the resulting depreciation of the dollar would make inflation shoot up. Others blame the inflation that ensued on the first oil shock – even though that happened 26 months later!
But if one looks closely, inflation had been rising since the mid 60´s mostly as the result of a strong belief in the stability of the Phillips Curve trade-off between inflation (politically acceptable since still relatively low) and unemployment (the reduction of which brought out the desired votes).
Figure 2 shows what was really driving inflation. Since the early 1960´s base money was trending up. As usual, especially if inflation has been low and stable as during 1960 – 1965, workers and firms, as a rule of thumb since not much will be gained from devoting resources to forecast inflation, will expect inflation to remain low i.e. πt+1= πt (inflation expected for next year will be the same as this year´s). But people learn and when money growth impacted on prices, it became a whole new ball game and people´s expectation formation adjusted to the new reality (this insight before the fact was one of the reasons Milton Friedman – and later Edmund Phelps – was awarded a Nobel).
The natural conclusion is that inflation is not a “victim” of the exchange rate or oil prices; quite the contrary, it was responsible both for the breakdown of the fixed (but adjustable) exchange rate system devised at Bretton Woods and for the oil price jump on October 1973! And obviously, behind inflation, is the economic and monetary policy of the government and the Central Bank. But every pol or official would rather blame it on someone else (unions, oligopolies, market speculators or oil producers).
“Recent history” nowadays is very different from what it was in 1986, so no one worries about the impact of currency moves on inflation in industrialized countries. In the case of the dollar what we hear on and of – following the large and persistent currency moves observed since the early 80´s – are worries about the “death” of the currency or the “end” of American economic power.
But in emerging markets in general and in Brazil in particular, the exchange rate – inflation/disinflation nexus is very much alive. Antonio Carlos Lemgruber´s post of August 28 (The Role of Monetary Policy under an Inflation Target Model: Brazil 2008) reflects the majority opinion. There we read that: “In the case of Brazil, for example, nobody doubts that a 100% – 50% when the exchange rate is defined as units of domestic currency per US dollar – appreciation against the dollar since 2003 was a major factor to keep inflation low…And this appreciation was mainly due to the high level of nominal interest rates against other currencies…”.
I, for one, doubt very much. Take figure 3, which shows that different currencies appreciated in tandem against the dollar since early 2002 (despite very different interest rate environments). The sole exception is the Brazilian real that between April and October 2002 depreciated 70%!
After the “fear of Lula” moment passed, the real had a lot of “catching up” to do! The anxiety about exchange rate moves in Brazil come from what can be observed in figure 4.
There we see that the 2002 rise in inflation followed the large depreciation, just as the fall in inflation after 2003 came on the heels of a strong appreciation. Figure 5 shows money growth and inflation over the same period.
The Granger causality issue, which has to do with precedence reminds me of Christopher Sim´s quip about the rooster and sunrise (if someone kills the rooster it doesn’t follow that the sun will not rise). Anyway, the point here is that the exchange rate depreciation in 2002 reacts immediately to the “bad news” about the next party in power. With that the government becomes restricted in the issue of debt and money (inflationary) financing is the alternative. Inflation expectations jump and so does the exchange rate. Only later do prices reflect this state of affairs.
It is interesting to note that in late 2000 and extending into 2001, the exchange rate also depreciated markedly, this time around due to contagion from the Argentine crisis. Despite all that, inflation remained stable. Figure 5 shows that monetary growth was also level.
When the “fear of Lula” subsided in late 2002, the government was again able to issue debt, money growth came down strongly and so did inflation expectations, the exchange rate and inflation.
Again, inflation is not the “victim” or “prize” of an exchange rate change, but quite the opposite. And once more, behind inflation we detect the behavior of economic policy, in this case initially constrained by the doubts flowing from the upcoming (October 2002) presidential election.
Now the newspapers are full of comments about the inevitable inflation that will follow the strengthening of the dollar relative to most currencies that began as July turned into August. If that is an unquestionable truth for Brazil it must also be so for all the other countries; but as goes the song: “It ain’t necessarily so”.